clear
import dataframe.*;

%task 1
%-------------------------------------------------------------------
%make chip--clinical id mapping file
ChipInfo = TextArray('listname','\t', 'NaN', 0);
%the first column stores the chip file name
ChipNames = ChipInfo.columnm(1);
%parse the chip file names into useful infomation
ChipInfo = TextArray(ChipNames,'_.');
%the forth column gives the unique id similar to clinnames
ChipID = ChipInfo.columnm(2);
ClinInfo = TextArray('Lung_fixed.clinical.txt','\t', 'NaN', 1);
ClinNames = deblank(cellstr(char(ClinInfo.columnm('Genome #'))));
if (length(ClinNames) ~= length(ChipNames)) 
    error('The number of chips do not match the number of samples');
end 
columns = cell(2,1);
columns{1} = ChipNames;
columns{2} = ClinNames;
headers = {'chip_file_name', 'clin_id'};
Map = TextArray(columns, headers);
Map.set('\t','NA',1);
Map.save('lung_chip_clin_map.txt');
%-----------------------------------------------------------------

%task 2
%-----------------------------------------------------------------
%make normalized clinical file
clin = TextArray('Lung_fixed.clinical.txt','\t', 'NaN', 1);
clin.setKeys(clin.columnm('Genome #'));
clin.saveAll('temp.txt');
clin = Tools.normaliseClinKeys('lung_chip_clin_map.txt','temp.txt',1,2);
clin.save('Lung_fixed.clinical.dfm',true,false,true);

%task 3
%--------------------------------------------------------------------
%make standard rma text array
y = Tools.makeTextArray('lung_rma.dat','listname','h133+.genedescriptors');
y.save('lung_rma.dfm',true,false,true);
%---------------------------------------------------------------------

%task 4
%--------------------------------------------------------------------
%use code table 
d = TextArray('Lung_fixed.clinical.dfm','tab','NaN',1);
columns = cell(1,1);
columns{1} = 'Race';
columns{2} = 'Sex'; 
columns{3} = 'Stg';
columns{4} = 'Diff';
columns{5} = 'CellType';
columns{6} = 'VascInv';
columns{7} = 'LymphInv';
columns{8} = 'LN Inv';
columns{9} = 'Pleural Inv';
codetable = d.createCodeTablem(columns,true);
codetable.save('Lung.Clinical.CodeTable.dfm',true,false,true); % then edit the codetable file as desired

codetable = TextArray('Duke.Clinical.CodeTable.dfm','tab','NaN',true);
%% That sets up the code table, and now change the clinical data in d to
%% numeric using this codetable 
%
d.applyCodeTable(codetable); 
d.saveAll('BreastCancer.Duke101.Clinical.Dec.20.04.dfm');  % saves as text file
d1 = d.getDataArray;
d1.saveAll('d1.dfm');
%--------------------------------------------------------------------------

%task 6
%---------------------------------------------------------------------
%merge clinical and gene expression data
Tools.mergeCG('Lung_fixed.clinical.dfm','lung_rma.dfm','lung_combined.dfm');
%---------------------------------------------------------------------

%task 6
%--------------------------------------------------------------------
%use code table 
d = TextArray('clin_s.dfm','tab','NaN',1);
columns = cell(1,1);
columns{1} = 'Study';
columns{2} = 'Path Stg'; 
codetable = d.createCodeTablem(columns,true);
codetable.save('Duke.Clinical.CodeTable.dfm',true,false,true); % then edit the codetable file as desired
codetable = TextArray('Duke.Clinical.CodeTable.dfm','tab','NaN',true);
%% That sets up the code table, and now change the clinical data in d to
%% numeric using this codetable 
%
d.applyCodeTable(codetable); 
d.saveAll('BreastCancer.Duke101.Clinical.Dec.20.04.dfm');  % saves as text file
d1 = d.getDataArray;
d1.saveAll('d1.dfm');
%--------------------------------------------------------------------------